a hobby class on microcontrollers
in [Embedded]
|
Prev: STUDIOT::Re: low-power-cons, low-cost, uP+DSP combo for H.264 videoencoding & control/monitor application
Next: Multimeter Clock – Styled after the Simpson 260 Multimeter
From: Jon Kirwan on 28 Jun 2010 20:12 On Tue, 29 Jun 2010 00:01:03 +0200, Hans-Bernhard Br�ker <HBBroeker(a)t-online.de> wrote: >Am 28.06.2010 20:44, schrieb Jon Kirwan: > >> I wanted DIP because I want the _actual_ micro piece used in the >> final project to cost very little, > >I don't think that's a valid line of reasoning. Despite your assurances contrary-wise, it is. >I seriously doubt that >the part being DIP has a guaranteed positived correlation to its >cheapness. No.. no.. that's not what I meant. I completely agree with you that DIP is __more__ expensive. They are more expensive to make. I've no doubt of that. Plus, if making in quantity helps cut costs further, then DIP will be more expensive for that reason, too. The issue isn't the difference between say $1.20 for DIP and $1.00 for SMT. The issue is being able to use cheap, easy to understand tools in wiring up. A DIP socket is very cheap. Even a ZIF socket for DIP is way cheaper than a ZIF socket for some other beast. And very inexpensive soldering irons and ham-fisted neophytes using them can be successful (enough times to not be completely frustrated, anyway) soldering to them. The 0.1" spacing is a minimum. I'd rather the parts used 1/4" spacing between pins. But that's not in the cards. The main thing is being able to hold something in one's hand, not lose it when you sneeze, and basically imagine these are 75 year olds or 15 year olds doing this. That is, if I allow them to use a soldering iron at all. Which I probably won't. Chances are, some of this will be done with Global Specialties prototyping boards and they will need to insert DIP parts into it and use jumper wires to connect things up. I _may_ introduce some of them to wire-wrap. Also, Linnix's board is all SMT and it costs (let's say) $10 each. If a student needs two of them, they pay $20. But in the case where a processor is available separately, then it might only be $1 or $2 or $3, instead, since a whole board doesn't have to come with it. And Linnix's board doesn't have a socket on it so that's a problem. The TI board comes with a socket. So there are a number of considerations included when I say "cheaper than." >Not these days, any more. > >I imagine the overhead in bond wire length, plastic material and pins >must be a major factor in the total price for microcontrollers by now >--- those with few enough pins to be eligible for putting in DIPs in the >first place, anyway. Not to mention economy of scale. I.e. even the >same chip in a different housing should generally be cheaper than the >DIP version. > >Let's face it, guys: cheaply socketable DIP is a dead parrot. The only >reason it's still sitting upright on its perch is that its feet have >been nailed to it. Just you aren't getting what I was talking about, that's all. I probably wrote more poorly than I should have. Jon
From: Jon Kirwan on 28 Jun 2010 20:19 On Mon, 28 Jun 2010 15:56:55 -0700 (PDT), steve <bungalow_steve(a)yahoo.com> wrote: ><snip> >I think you have conflicting requirements, something easy to use is >going to require big knobs, cables, big connectors and hopefully >solderless === big money. A small flat pcb board with nothing but >holes in it (TI 4.30 etc) is too intimidating for your audience. The reality will be in the details. And where all this winds up, I don't know right now. These kits, were you able to actually open them and look inside, come with large and small switches, knobs, wheels, gears and gear boxes, dc motors, 40kHz emitters and receivers, LEDs, and a host of other useful items. I've already got enough stuff to keep the costs of a successful project low and I don't need ANY MORE parts, at all, to do that. So I know the actual cost here. It may cost more, once I run out of this stuff. But with hundreds of boxes like this, that may take a bit of time and by then I will know more and can work on the exact supplies I will need. On the last point, the intimidation factor might be there _before_ they get into the classroom. I can't control their minds before that point in time. So that is why I mentioned that I need to figure out how to "sizzle" the class. But once they are there, it is MY JOB to take away any feelings of intimidation. That's why I get paid the big bucks ($0) to do this. If I haven't done my job well, then your point is valid. But if I succeed as I hope I may, then I'm more worried about specific details I need to smooth over and resolve than about this. But I am very glad for your contributing thoughts, just the same. I need to hear objections and consider them well. Jon
From: Rob Gaddi on 28 Jun 2010 20:32 On 6/28/2010 5:19 PM, Jon Kirwan wrote: > On Mon, 28 Jun 2010 15:56:55 -0700 (PDT), steve > <bungalow_steve(a)yahoo.com> wrote: > >> <snip> >> I think you have conflicting requirements, something easy to use is >> going to require big knobs, cables, big connectors and hopefully >> solderless === big money. A small flat pcb board with nothing but >> holes in it (TI 4.30 etc) is too intimidating for your audience. > > The reality will be in the details. And where all this winds > up, I don't know right now. > > These kits, were you able to actually open them and look > inside, come with large and small switches, knobs, wheels, > gears and gear boxes, dc motors, 40kHz emitters and > receivers, LEDs, and a host of other useful items. I've > already got enough stuff to keep the costs of a successful > project low and I don't need ANY MORE parts, at all, to do > that. So I know the actual cost here. > > It may cost more, once I run out of this stuff. But with > hundreds of boxes like this, that may take a bit of time and > by then I will know more and can work on the exact supplies I > will need. > > On the last point, the intimidation factor might be there > _before_ they get into the classroom. I can't control their > minds before that point in time. So that is why I mentioned > that I need to figure out how to "sizzle" the class. But > once they are there, it is MY JOB to take away any feelings > of intimidation. That's why I get paid the big bucks ($0) to > do this. If I haven't done my job well, then your point is > valid. But if I succeed as I hope I may, then I'm more > worried about specific details I need to smooth over and > resolve than about this. > > But I am very glad for your contributing thoughts, just the > same. I need to hear objections and consider them well. > > Jon Have you checked out some of the MSP430 dev kits? They've got this new widget http://e2e.ti.com/support/microcontrollers/msp43016-bit_ultra-low_power_mcus/f/165/p/53135/188234.aspx#188234 that they're claiming is $4.30/ea. Then there's their $20 eZ430 kits. The $20 gets you an emulator with a detachable target board, then you can get more target boards at $10/3pc. -- Rob Gaddi, Highland Technology Email address is currently out of order
From: David Brown on 29 Jun 2010 03:17 On 29/06/2010 01:43, Gary Peek wrote: > Hans-Bernhard Br�ker wrote: >> Let's face it, guys: cheaply socketable DIP is a dead parrot. The only >> reason it's still sitting upright on its perch is that its feet have >> been nailed to it. > > I'm not sure what "dead parrot" means, <http://en.wikipedia.org/wiki/Dead_Parrot> <http://www.youtube.com/results?search_query=dead+parrot&aq=f> If you are not a Monty Python fan, it might be hard to appreciate the humour in it, but basically a "dead parrot" is something that is dead, but some people claim it is actually still alive. > but for a hobby board (or any > board that someone will be "experimenting" with) having the chips in > sockets is an advantage come repair time. > Unless you have nothing much on the board, or everything else is socketed, then there are few advantages in having the microcontroller socketed. It is helpful to have reasonably large pitch, so that you can solder the chip, but I don't understand the big desire for DIPs for a microcontroller.
From: Jon Kirwan on 29 Jun 2010 04:09
On Mon, 28 Jun 2010 17:32:24 -0700, Rob Gaddi <rgaddi(a)technologyhighland.com> wrote: >Have you checked out some of the MSP430 dev kits? They've got this new >widget >http://e2e.ti.com/support/microcontrollers/msp43016-bit_ultra-low_power_mcus/f/165/p/53135/188234.aspx#188234 >that they're claiming is $4.30/ea. Then there's their $20 eZ430 kits. >The $20 gets you an emulator with a detachable target board, then you >can get more target boards at $10/3pc. Yes, if you look over my posts in this thread you will see that I've already mentioned the one for $4.30 and I may have also mentioned the other one as well. In any case, I've posted up some tutorials that use the device on the Yahoo MSP430 group, as well. So yes, I'm well aware of them (and the extra target 2012 boards, too, as I've bought many, many dozens of those, as well.) Thanks, Jon |